International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

A Gis Based Uhf Radio Wave Propagation Model for Area Within 25km Radius From OSRC Transmitting

K. L. Omolaye, P. O. Omolaye, G. A. Akpakwu, Dept of Geographical Information Dept. of Electrical and Electronics Dept of Electrical and Electronics System and Remote Sensing, Engineering, Engineering, Federal University of Technology, University of Agriculture , University of Agriculture, Makurdi, Akure, Nigeria. Makurdi, Nigeria. Makurdi, Nigeria.

Abstract- For effective network coverage across different terrain The wireless communication has drawn attention as a and landcover, accurate prediction model is required. Radio wave result of drastic change in technology and the demand for the models are essential tools for predicting wave propagation quality products and services over the years. From the characteristics in a particular environment. This paper investigates transmitter, takes different propagation path signal strength using Geospatial Technology which was applied to to the receiver and it depends on the interaction with UHF radio wave propagation within 25km radius from Ondo State Radiovision Corporation Transmitting Antenna. Five major routes interfering object along the path of propagation. In radio wave within the study area were considered and signal strength propagation, there are interactions between waves and measurements were carried out at interval of 1km across different environment attenuates the signal level which causes land cover type and terrain variations. Three empirical models and finally limits coverage area. This is simply because (Okumura-Hata, COST-Hata and ) were considered for propagation characteristics of radio wave such as fading, path the study due to their system parameter. The path loss of the loss and attenuation do not only depend itself on frequency, empirical model was compared with measured path loss and the path distance, and characteristics scatter distance, but also on mean deviation errors of the models were computed. Consequently, scatter angle that depends on what is causing obstruction to Okumura-Hata’s model is suitable for prediction along the five IJERTthe propagated wave. Propagation models are classified into routes. A modified Okumura- was developed for pathIJERT loss prediction in 25km radius away from the transmitting antenna three main groups: deterministic, statistical and empirical which is applicable for network planning and design. models [3]. Deterministic models make use of the physical laws which determine radio wave propagation mechanisms for Keyword: Geospatial, Ondo, Empirical models, path loss, a particular location. Stochastic (statistical) models, on the Okumura-Hata, COST-Hata and Egli model other hand, require the least information about the environment but provide the least accuracy. Empirical models 1. INTRODUCTION are based on extensive measurements and mainly give prediction of path loss. They are more often used in practice Radio waves are part of the broad electromagnetic than statistical and deterministic propagation models, because spectrum that extends from the very low frequencies which are of low cost and model simplicity with acceptable accuracy''. produced by electric power facilities up to the extremely high Examples of wireless models include the Okumura-Hata frequencies of cosmic rays which travel from a source into the model, COST 231-Hata model, COST 231-Walfisch-Ikegami, surrounding space at the speed of light, approximately Friis model, 2D and 3D ray tracing, SUI model, Cluster Factor 3.0x108 m/s when in free space. The electromagnetic model, Walfisch-Bertoni model, Eglic model, Erecg Model, spectrum extends as low as sub-hertz frequencies to 30 000 COST 123 model etc. Each of these model are useful for GHz. They are divided into several bands by frequency in specific environment and the accuracy of any of the model which FM and TV broadcast is a good example with a very depends on the parameter required by the model and the high frequency band from 30 to 300MHz [1]. In radio environment. There is no universal model for all kind of broadcasting, the energy that is emitted from an antenna takes propagation situation. All propagation models are based on the different paths before it reach the receiver and radio wave terrain parameters (rural, urban and suburban area) and system pathway depends on many factors, some of which include: parameter like antenna height, frequency (Abhayawardhana, et frequency, antenna type and height, atmospheric conditions, al). All computations in radio communication are based on the land use variation and terrain [2]. use of radio propagation prediction models. It is also reported that, Geographical Information System (GIS) as a tool in

IJERTV3IS050893 www.ijert.org 1784 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

partner with the field of remote sensing has become an important component for modeling radio wave prediction [4]. Until recently, empirical propagation prediction South, Ondo East, Idanre, Ifedore, Okeigbo/Ile Oluji Local model seemed sufficient. However, more efficient prediction Government), Ekiti State (Ekiti South West, Ikere and models are required in transmitting data (voice and non voice) Ise/Orun Local Government) and Oriade Local Government from the antenna to the receiver. These propagation models in Osun State. The study area was found to have minimum are usually based on the computation of the physical elevation value 204m above the mean sea level and maximum interaction of radio waves and the environment. Thus, more elevation value of 1065m above the mean sea level. detailed geo-databases of the environment are required, especially in urban environments where most users are located [5]. In broadcast frequency planning, GIS is relevant in base station programming, base station position selection, field strength prediction, coverage analysis, interference analysis, frequency distribute and so on. For accurate prediction of radio wave, interference analysis and coverage calculation demand detail location of spatial database of the predicted area. The evaluation of the wireless technology grows; the industry development trend uses geographic information system to strengthen TV broadcasting, coverage, network management and construction [6]. The inability of radio wave to penetrate or bend around geographic features which result in non line of sight can be easily modeled in geographical information system. GIS have ability to model the Figure 1: The map of the Study Area communication sheds (commshed) which is also known as the

viewshed. Having the LOS properties of the wavelength, the III. RESEARCH METHODOLOGY viewshed shows the zone where there will be signal [7].

a. Data and their Sources Ondo State Radio vision Corporation propagates in Table 1: Data and their sources UHF band. The UHF band goes from 300 MHz to 3GHZ which is the ideal choice for ground to air communication. Data Year Source Relevance UHF has a wide band width which propagates principally on Google image 2007 Google For route design and line of sight from the transmitter to the receiver. The effects of earth navigation IJERTASTER 2011 USGS For digital elevation local area topography and conditions in the lower atmosphereIJERT model and spot height mostly govern UHF propagation. For communication to take Field 2014 Self For signal strength place the transmitting and receiving antennas must have a measurement measurement fairly unobstructed path between them, hence the term line-of- Antenna 2013 OSRC For path loss sight can be established [8]. Radio wave is a function of description calculation Landsat 2014 USGS For landuse/cover frequency of propagation, lower radio frequencies, such as classification AM/FM radio, have lower wavelengths which allow them to

penetrate geographic features such as vegetation, building wall b. Observation Instruments and others. As the wavelength decreases, the frequency increases which make more influence by geographic feature in 1. Garmin 76c handheld GPS was used basically to determine the point of observation and the transmitting the environment [9]. It is stated that the quality of signal strength is extensively blocked or degraded by nature and man antenna. to the relatively short wavelength propagation [4]. Geographic features hinder the propagation of wave which can be easily model Invisibility theory i.e. line of sight theory [10].

II. STUDY AREA AND CHARACTERSTIC

The study area covers 25km from Ondo State Radiovision Corporation Antenna which lies within latitudes Fig 2.1: Global Positioning System (GPS) used for direct observation 6° 30’ 00” and 8° 00’ 00” and longitudes 4° 30’ 00” and 5° 30′ 00'' which cut cross three states; Ondo, Osun and Ekiti States. 2. Professional TV signal field strength meter type Within the three states, the study area covers ten Local UNAOHM model EP742A was used measure signal field Government across the state; Ondo State (Akure North, Akure strength which was carried out round the selected route.

IJERTV3IS050893 www.ijert.org 1785 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

Landcover Classification: The land use/cover of the study area was obtained from Lansat 8 imagery downloaded from USGS and subjected to unsupervised classification using ENVI 4.5 which gave a general idea of the land use/cover type of the area. Bands 543 were combined to obtain a colour composite that was used for the unsupervised classification. . Research showed that the different land cover/use type affect signal strength [11], [12]. Fig 2.2: Field strength Meter used for direct observation

3. A Yagi array receiving antenna covering both VHF and UHF frequency bands was used for measurements. This was mounted on support about eight metre (8m).

Fig 2.3: Yagi Antenna used for direct observation

IV. METHODS Figure 3.2: Land cover classification of the study area The method employed for this research involves measuring signal strength which relies on GIS and remote Digital Elevation Model: For this study, ASTER DEM at 30m sensing techniques for measurement and visualization. This resolution was used for the study area which was further research was based on three characteristic which are terrain, reclassified into low and high. Spot height across the study land cover type and distance from the transmitting antenna. area was generated using TIN node tools in ArcGIS. Elevation The method adopted for this research is as follows: value and the corresponding signal strength was noted at every IJERTIJERT1km interval from the transmitter. Driving Route Test Design: The GPS coordinate of the transmitting antenna was obtained. This was plotted in ArcGIS environment and 25km ring buffering around the transmitting antenna was carried out at 1km interval. Points where the buffer crosses the route were recorded. This coordinates were traced to the ground for signal observation.

Figure 3.1: Driving route drive Figure 3.3: Digital Elevation Model of the study area

IJERTV3IS050893 www.ijert.org 1786 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

Signal Strength Measurement: The signal strength V. RESULT AND DISCUSSION measurement was carried out at frequency 487.25MHz based on the coordinate obtained during driving route design. a. Result of Television Signal Strength Measurement Coordinate information from ArcGIS software for every point of observation was used after carrying a ring buffer operation The following tables show the field strength from the transmitting antenna at every 1km to 25km. Signal measured across each of the route at 1km interval from the strength across terrain variation was measured and recorded at transmitter at frequency 487.25MHz. The location data (x,y) every route of the observation point was considered for signal measurement and there was direct line of sight observed for Signal Strength Measurement with Expected Signal measurement in this study. Below is table and corresponding Strength Value: The signal strength measurement was carried graph. out based on the above route design. Measurement was carried out in different defined route land use type and terrain classification in order to predict signal strength. Distance and line of sight (LOS) from the transmitter to the receiver were considered. Field signal strength data was acquired within ten days. It was carried out in late January with little or no atmospheric interference. The measured signal was compared to theoretical value. Equation 1 was used to calculate theoretical values as obtained.

푉 30푃 퐺 퐸 = 푡 푡 (1) 푚 푑(퐿푂푆)

where Pt is the transmitted power in (W), Gt is the gain of the transmitter, and d is the LOS distance in (m). Figure 4.1a: Graph of field strength measurement in Ilesa

Signal Strength Measurement with Existing Models: The field measurement was carried out in the study area and compared with empirical models that are developed to predict signal strength. The empirical model used in this study are Okumura Hata, Egli and COST 123 model which was compared with field measurement in the study area in other toIJERT IJERT develop a based model for the study area. Equations 2, 3 and 4 were used to determine the corresponding path loss for each measurement taken.

88 푃ℎ ℎ 푑푁−2 퐸 = 푟 푡 ℎ (2) ℎ 휆푑2

푃 퐺 퐺 푃 = 푡 푡 푟 (3) 푟 푃퐿 Figure 4.1b: Graphic of field strength measurement in Ijare route 푃 퐺 퐺 ℎ2ℎ2 푃 = 푡 푡 푟 푡 푟 (4) 푟 푑4 Where Pr is in watts, hr, ht is in metre, dh, d are in Km and f is the frequency in MHz, [13].

Figure 4.1c: Graphic of Signal strength measurement in Ado-Ekiti route

IJERTV3IS050893 www.ijert.org 1787 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

Figure 4.2b: Line graph showing the relationship between the theoretical Figure 4.1d: Graphic of signal strength measurement of Ondo route measurement and field strength measurement in the Ilesa route

Figure 4.1e: Graphic of field strength measurement in Idanre Figure 4.2c: Line graph showing the relationship between the theoretical measurement and field strength measurement in the Ondo route From the television signal strength measured across the five route, it was observed that their are direct line of sight between the transmitter and the receiver but the signal strengthIJERT IJERT reduces as reciever is moved away from the transmitter. The reduction in signal strength is not uniform which is as a result of attenuation caused by different land cover/use, atmospheric refraction and other.

b. Result of Measured Value Compared with Theoretical Value

The following table shows the measured against the theoretical measurement

Figure 4.2d: Line graph showing the relationship between the theoretical measurement and field strength measurement in the Ijare route

Figure 4.2a: Line graph showing the relationship between the theoretical measurement and field strength measurement in the Idanre route

IJERTV3IS050893 www.ijert.org 1788 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

d. Result of Measurement obtained compared to Empirical Model for each route

Fig 4.2e: Line graph showing the relationship between the theoretical measurement and field strength measurement in the Ado-Ekiti route

c. Result of Comparism for Measured Signal and the Figure 4.3b: Path losses in Ilesha route with existing model Empirical Model Path Loss Calculation

Okumura-Hata model, COST-Hata model and Egli model were used to predict the path losses along the five routes and the results are as shown in Tables 3. Path loss was computed by using the reference station specification from the model used: Okumura, COST-Hata and Egli model was reduce to the equation 5, 6and 7 respectively

PL = 119.892 + 26.44logd (5)

P = 93.988 + 63.36logd (6) L 2 Figure 4.3c: Path loss in Ondo route with existing model PL = 68.01/d (7) IJERTIJERT

Figure 4.3d: Path loss in Idanre route with existing model Figure 4.3a: Path loss predictions with existing model

From the above diagram, Path loss for each model was calculated at each LOS. The result shows that Egli model assumes that there is obstruction such as buildings and hilly area that is why the path loss is low which is not true while Okumura-Hata and COST 231 shows better prediction of signal strength. The overlap that occurs between the two models shows that Okumura Hata model was modified to develop COST 231 model.

IJERTV3IS050893 www.ijert.org 1789 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

Table 3: Comparison of Path Loss of Empirical Models with Measurements for Ondo Route.

Model Okumura- COST- Egli Hata Hata

Path Loss Mean Error 50.01 61.31 93.87 (dB)

Table 4: Comparison of Path Loss of Empirical Models with Measurements for Idanre Route.

Model Okumura- COST- Egli Figure 4.3e: Path loss in Ijare route with existing model Hata Hata

Path Loss Mean Error 48.61 59.91 95.27 (dB)

Table 5: Comparison of Path Loss of Empirical Models with Measurements for Ijare Route.

Model Okumura- COST- Egli Hata Hata

Path Loss Mean Error 50.05 61.34 93.83 (dB)

Figure 4.3f: Path Loss in Ado-Ekiti route with existing model

From the above path loss across the five route Table 6: Comparison of Path Loss of Empirical Models with Measurements considered in this research, it was observed that Egli model for Ekiti Route. was out rightly ineffective for the predicted area. The model shows obstruction along the path of observations. At each of Model Okumura- COST- Egli IJERT Hata Hata the distance where measurement was observed, there is directIJERT line of sight from the receiver to antenna. For each of the Path Loss Mean Error 46.84 58.14 96.94 route, is better than COST 231 model in this (dB) study area based on closeness of path loss to the measured value along each of the paths.

e. Comparism of Empirical Models with Measurements f. Discussion: The corresponding error statistics in terms of the mean It was observed from the analysis that out of the three prediction error are shown in the Tables 2, 3, 4, 5 and 6 were models considered, Okumura-Hata model has minimum error gotten from the above graphs for each of the route. of 52.08dB along Ilesa route, 50.02dB along Ondo route, 48.62dB along Idanre route, 50.05dB along Ijare route and Table 2: Comparison of Path Loss of Empirical Models with Measurements 46.85dB along Ekiti route. Also, the measurements taken were for Ilesa Route. compared with the theoretical estimations for validation which shows the theoretical measurements as obtained from equation Model Okumura- COST- Egli Hata Hata (1) with respect to the LOS distance and the measurements taken in the 25km radius for the five chosen routes around the Path Loss Mean Error 52.08 63.37 90.08 antenna. With the disparity shown between these measured (dB) values along the five chosen routes, it could be deduced that the propagation here suffered attenuation caused by physical environment, atmospheric refraction; and the neutral atmosphere induce propagation delays. In the neutral

atmosphere, delays are induced by refractivity of gases, hydrometeors, and other particulates, depending on their permittivity and concentration, and forward scattering from

IJERTV3IS050893 www.ijert.org 1790 International Journal of Engineering Research & Technology (IJERT) ISSN: 2278-0181 Vol. 3 Issue 5, May - 2014

hydrometeors and other particulates. Changes in temperature, prediction at 25km radius around the transmitting antenna moisture, and pressure in the atmospheric column cause a which will assist engineers in network planning and design. change in atmospheric density, which in turn causes variations in the intensity of waves in both the vertical and horizontal. Reflection and diffraction caused by obstruction (buildings, REFERENCES mountains with different elevations among others) and the effect of tree density with foliages along the paths chosen. The [1] Flickenger, R. (2007). Wireless Networking in the Developing presence of vegetation produces a constant loss, independent World (2nd ed.). LLC: Hacker Friendly. [2] Barringer, M. H., and Springer, K. D. (1992). Radio wave of distance between communication terminals that are spaced propagation ( 8th. ed.). Washington, DC: NAB. 1 km or more apart. Since the density of foliage and the [3] V.S.Abhayawardhana, I.J.Wassell, D.Crosby, M.P.Sellars, and heights of trees are not uniformly distributed in the area M.G.Brown. (2005). Comparison of empirical propagation path (forested environment) and corresponding error statistics in models for fixed wireless access system. VTC 2005-Spring.IEEE 61st , vol. 1, pp.73-77. terms of the mean prediction error. It could be observed [4] Rose, S. (2001.). The Effect of Digital Elevation Model Resolution clearly that all the three empirical models under predicted the on Wave Propagation Predictions at 24 GHz. Virginia Tech,: MS path loss with COST-Hata model and most especially Egli thesis in Geography. model grossly under predicted the path loss (i.e. Okumura- [5] Jean-Frederic, W., and Rizk, K. (2003). Radiowave propagation, building databases, and GIS: anything in common? A radio Hata model has minimum error in the prediction). engineer's viewpoint. Environment and Planning B , 30, 767-787. [6] Bing, L., Chang, X., and Ruoyi, w. (nd). TV Broadcasting Consequently, Okumura-Hata’s model is suitable for Frequency Programming based on GIS. Liaoning. prediction along the five routes, while Okumura-Hata’s model [7] Dodd, H. M. (2001). “The Validity of Using a Geographic Information System’s Viewshed Function as a Predictor for the is suitable for path loss prediction along those paths taken. Reception of Line-of-Sight Radio Waves. Virginia Tech: MS thesis in Geography. VI. CONCLUSION [8] Harris. (2000). Radio Communications in the Digital Age (Vol. vol 2: VHF/UHF Technology). USA: Harris Corporation. In this research work, three empirical propagation [9] Weisman, C. J. (2000 ). The Essential Guide to RF and Wireless. New Jersey: Prentice Hall. models: Okumura-Hata model, COST-Hata model, and Egli [10] Mardeni, R., and Kwan, K. F. (2010). Optimization of Hata model were used for path loss computation along the selected propagation prediction model in suburban in Malaysia. Progress In five routes. Along the five routes chosen it was observed that Electromagnetics Research C, , Vol. 13, 91{106,. COST231-Hata model and Egli model showed larger mean [11] Allsebrook, K., and Parsons, J. D. (1977). Mobile Communications Design Fundamentals. IEEE , 23, 312-322. path loss error i.e. they grossly under predicted the path loss, [12] Parson, J. D. (1983). Propagation Engineering in Wireless while Okumura-Hata model showed closer agreement with the Communications (2 ed.). New York: McGraw-Hill Inc. results measured with lower mean path loss error. Therefore, IJERTa IJERT[13] Mithal, G., and Mittal, R. (1964). Radio Engineering (1 ed.). New modified Okumura-Hata model was developed for path loss York: McGraw.

IJERTV3IS050893 www.ijert.org 1791